U.S. patent number 5,638,535 [Application Number 08/441,405] was granted by the patent office on 1997-06-10 for method and apparatus for providing flow control with lying for input/output operations in a computer system.
This patent grant is currently assigned to NVidia Corporation. Invention is credited to Chris A. Malachowsky, Curtis Priem, David S. H. Rosenthal.
United States Patent |
5,638,535 |
Rosenthal , et al. |
June 10, 1997 |
Method and apparatus for providing flow control with lying for
input/output operations in a computer system
Abstract
A flow control circuit for a computer system including a
first-in first-out buffer including a register for storing a value
indicating the number of stages of the FIFO which are available to
store data, circuitry for detecting whether an input/output device
is able to process data more rapidly than the FIFO is emptied, and
circuitry for providing an value greater than the number of stages
actually available for storage in the FIFO if the input/output
device is able to process data more rapidly than the FIFO is
emptied.
Inventors: |
Rosenthal; David S. H. (Palo
Alto, CA), Priem; Curtis (Fremont, CA), Malachowsky;
Chris A. (Santa Clara, CA) |
Assignee: |
NVidia Corporation (Sunnyvale,
CA)
|
Family
ID: |
23752748 |
Appl.
No.: |
08/441,405 |
Filed: |
May 15, 1995 |
Current U.S.
Class: |
711/165; 710/1;
710/34; 710/57 |
Current CPC
Class: |
G06F
5/06 (20130101) |
Current International
Class: |
G06F
5/06 (20060101); G06F 012/00 () |
Field of
Search: |
;395/877,849,854,821,823,876 |
References Cited
[Referenced By]
U.S. Patent Documents
Other References
BIST for ring-address SRAM-type FIFOs by Goor et al., 1994 IEEE
publication, pp. 112-118. 1994..
|
Primary Examiner: Treat; William M.
Assistant Examiner: Maung; Zarni
Attorney, Agent or Firm: King; Stephen L.
Claims
What is claimed is:
1. A flow control circuit in a computer system comprising:
a first-in first-out (FIFO) circuit having a plurality of stages,
the FIFO circuit functioning asynchronously with respect to a
central processing unit of the computer system,
means for informing a source of data of a number of stages of the
FIFO circuit which are available to store data,
storage means for storing data transferred to the FIFO buffer in
excess of the number of stages of the FIFO circuit which are
available to store data, and
means for providing a value greater than the number of stages of
the FIFO circuit which are available to store data to a source of
data.
2. A flow control circuit as claimed in claim 1 in which the means
for informing a source of data of a number of stages of the FIFO
circuit which are available to store data includes a register for
storing a value indicating the number of stages of the FIFO circuit
which are empty.
3. A flow control circuit as claimed in claim 1 in which the means
for informing a source of data of a number of stages of the FIFO
circuit which are available to store data includes a register for
storing a value indicating the number of stages of the FIFO circuit
which are empty, and
the means for providing a value greater than the number of stages
of the FIFO circuit which are available to store data to a source
of data comprises means for storing an indication that a device
which is to receive data typically clears data from the FIFO
circuit at the same rate or faster than the source of data is
capable of providing data.
4. A flow control circuit as claimed in claim 1 in which the means
for storing an indication includes an address translation circuit
for providing a physical address of a device which is to receive
data and the indication.
5. A flow control circuit as claimed in claim 1 further
comprising:
means for detecting when the FIFO circuit is full, and
means responding to an indication that the FIFO circuit is full for
transferring additional data to the storage means.
6. A flow control circuit as claimed in claim 1 in which the FIFO
circuit includes a plurality of individual FIFO buffers.
7. A flow control circuit as claimed in claim 6 further comprising
associated with each individual FIFO buffer:
means for detecting when the FIFO buffer is full,
means responding to an indication that the FIFO buffer is full for
transferring any additional data to the storage means.
8. A flow control circuit as claimed in claim 1 further
comprising:
means for processing data provided at an output stage of the FIFO
circuit,
means for processing data in the storage circuit,
means responding to an indication that data is being transferred to
the storage circuit for disabling the transfer of data to the FIFO
circuit,
means responding to an indication that data is being transferred to
the storage circuit for processing data stored in the FIFO circuit
until all data has been processed,
means for processing all data in the storage means until all data
in the storage means has been processed, and
means for enabling transfer of data to the FIFO circuit when all
data in the storage means has been processed.
9. A method for controlling the flow of data in a digital system
comprising a first-in first-out (FIFO) buffer having a plurality of
stages for transferring information serially and functioning
asynchronously with respect to a central processing unit of the
computer system, the method comprising the steps of:
detecting a first value indicating whether a device to which data
is to be transferred typically handles information faster than
information can be transferred to the FIFO buffer,
if the first value detected indicates that a device to which data
is to be transferred typically handles information faster than
information can be transferred to the FIFO buffer, detecting a
second value indicating an amount of space available for data in
the FIFO buffer,
if the first value detected indicates that a device to which data
is to be transferred typically handles information slower than
information can be transferred to the FIFO buffer, detecting a
third value indicating an amount greater than an amount of space
available for data in the FIFO buffer,
transferring a value detected to a source of information to control
the amount of information sent to the FIFO buffer,
transferring to the FIFO buffer an amount of data up to an amount
determine from the second or third value detected, and
detecting another value indicating an amount of space available for
data in the FIFO buffer before transferring additional data to the
FIFO buffer.
10. A method for controlling the flow of data in a digital system
as claimed in claim 9 further comprising the steps of:
detecting when the FIFO buffer is full, and
responding to an indication that the FIFO buffer if full by
transferring any additional data to storage means.
Description
BACKGROUND OF THE INVENTION
1. Field Of The Invention
This invention relates to computer systems, and more particularly,
to a method and apparatus for providing flow control of
input/output operations in computer systems.
2. History Of The Prior Art
Modern computer system are typically based on an architecture which
was first offered in the Digital Equipment Corporation (DEC) PDP 11
computer. One problem with this architecture as with earlier IBM
and CDC mainframe architectures is that writing directly to the
input/output devices of the system by an application program is
prohibited. Although this architecture allows all of the facilities
of the central processing unit to be used for input/output, it
requires that the operating system running on the central
processing unit attend to all of the input/output functions using
trusted code. This significantly slows any input/output operation
of the computer.
In contrast to earlier mainframe systems, in this architecture,
there is no process by which the input/output performance of the
system can be increased except by increasing the speed of the
central processing unit or the input/output bus. This is an
especial problem for programs which make heavy use of input
output/devices such as video and game programs which manipulate
graphics and high quality sound extensively.
In a modern computer, the central processing unit and the
input/output devices operate at different speeds. It can be very
inefficient for a modern central processing unit to wait until an
input/output write operation is complete before performing the next
operation which often has nothing to do with input/output. On the
other hand, a central processing unit has to wait for the result of
a read operation because it needs the result produced.
Since most central processing unit accesses to input/output devices
are write operations, the designers of systems and input/output
devices attempt to decouple the central processing unit and
input/output devices as far as write operations are concerned by
implementing write queues using first-in first-out (FIFO) write
buffers. These buffers may appear at various places in a particular
implementation: as a part of the central processing unit, as part
of a bridge chip, or as part of an input/output device.
One problem raised in systems using FIFO buffers is that an
input/output device and the buffers supplying it must accept all
information written to them over the input/output bus. Although
some input/output buses allow devices to "hold off" writes, that
is, delay the completion of the write operation until the device
has enough resources available to store the data, there is always a
limit to how long a write can be held off. If a write is held off
too long the data will be lost. In the limit, the input/output
device has no alternative but to store all data written to it. In a
system utilizing FIFO buffers for storage of this data at the
input/output device, the FIFO buffers must ultimately store the
data.
Since any practical input/output device will have limited FIFO
buffer storage for holding data written to it over the input/output
bus, any architecture for input/output devices must include some
technique for controlling the flow of data so that this storage is
not exhausted.
It is desirable to provide a means for providing flow control for a
computer system or similar system utilizing FIFO buffers to receive
data so that the operation may proceed as rapidly as possible
without loss of data. A new input/output architecture which allows
input/output operations to proceed at a faster rate by allowing
application programs to write directly to input/output devices used
with advanced multi-tasking operating systems has now been
designed. One of the features of this system is the use of a write
buffering arrangement including FIFO buffers. These FIFO buffers
include flow control circuitry including registers which store a
value indicating the amount of free space remaining in the FIFO
buffer. By ascertaining this amount before data is transferred to a
FIFO buffers and sending no more than that amount, no data is lost
and transmission speed is maintained. However, such a system
requires that the value indicating the amount of space available in
a FIFO buffer be transferred to the system processor continuously
during operation of the system. This places a substantial overhead
on the operation of the system.
In many situations the FIFO buffers are able to transfer data much
more rapidly than a processor can transfer data to the FIFO buffer.
Consequently, a FIFO buffer may be able to handle much more data
than indicated by the value of empty space in the FIFO buffer.
It is desirable to be able to handle data available at the FIFO
buffer without slowing the operation of the computer by constantly
reading a value in a flow control register yet still be able to
handle situations in which data is being processed slowly through
the FIFO buffers.
SUMMARY OF THE INVENTION
It is, therefore, an object of the present invention to provide a
means for accomplishing flow control of data being written by the
central processing unit on the input/output bus of a computer
system.
It is another object of the present invention to provide an
improved flow control system for an architecture which utilizes
FIFO buffers for buffering write operations.
These and other objects of the present invention are realized in a
system which uses an arrangement of parallel FIFO buffers in which
each FIFO buffer accepts data from only one application program. In
order to assure that no data written to a FIFO by an application
program will overflow the FIFO, each FIFO includes a flow control
register which must be read by the central processing unit running
the application before writing data to an input/output device. The
register stores a value which indicates the amount of space
available in the FIFO to which data may be written. Reading this
register tells the application program how much data may be written
without running the risk of overflowing the data storage area which
the input/output device has available. Provided that no application
program ever writes more than the data for which it has permission,
no data will be lost. In one embodiment, the arrangement includes
circuitry for detecting whether an input/output device is able to
process data more rapidly than the FIFO is emptied, and circuitry
for providing an value greater than the number of stages actually
available for storage in the FIFO if the input/output device is
able to process data more rapidly than the FIFO is emptied. If too
much data is furnished, circuitry is provided for processing the
overflow.
These and other objects and features of the invention will be
better understood by reference to the detailed description which
follows taken together with the drawings in which like elements are
referred to by like designations throughout the several views.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 is a block diagram of a computer system providing facilities
by which a central processing unit may write directly to
input/output devices.
FIG. 2 in a diagram illustrating the operation of software in the
architecture of the present invention.
FIG. 3 is a block diagram of one embodiment of input/output
circuitry used in a personal computer system such as that
illustrated in FIG. 1 designed in accordance with the present
invention.
FIG. 4 illustrates the address and data bits utilized in one
embodiment of the invention.
FIG. 5 is an illustration of entries in a translation table used in
accordance with the invention.
FIG. 6 is a block diagram of another embodiment of input/output
circuitry used in a personal computer in accordance with the
present invention.
FIG. 7 is a block diagram illustrating in more detail specific
portions of the circuitry shown in FIG. 6.
NOTATION AND NOMEMCLATURE
Some portions of the detailed descriptions which follow are
presented in terms of symbolic representations of operations on
data bits within a computer memory. These descriptions and
representations are the means used by those skilled in the data
processing arts to most effectively convey the substance of their
work to others skilled in the art. The operations are those
requiring physical manipulations of physical quantifies. Usually,
though not necessarily, these quantities take the form of
electrical or magnetic signals capable of being stored,
transferred, combined, compared, and otherwise manipulated. It has
proven convenient at times, principally for reasons of common
usage, to refer to these signals as bits, values, elements,
symbols, characters, terms, numbers, or the like. It should be
borne in mind, however, that all of these and similar terms are to
be associated with the appropriate physical quantities and are
merely convenient labels applied to these quantities.
Further, the manipulations performed are often referred to in
terms, such as adding or comparing, which are commonly associated
with mental operations performed by a human operator. No such
capability of a human operator is necessary or desirable in most
cases in any of the operations described herein which form part of
the present invention; the operations are machine operations.
Useful machines for performing the operations of the present
invention include general purpose digital computers or other
similar devices. In all cases the distinction between the method
operations in operating a computer and the method of computation
itself should be borne in mind. The present invention relates to a
method and apparatus for operating a computer in processing
electrical or other (e.g. mechanical, chemical) physical signals to
generate other desired physical signals.
DETAILED DESCRIPTION
FIG. 1 is a block diagram of a computer system 22 which has been
devised to overcome the problems of the prior art. The system 22
provides a new input/output architecture which cooperates with
other components of present systems based on the PDP11
architecture, runs legacy code for those systems, yet is able to
drastically increase the speed of input/output operations for new
application programs. In order to accomplish this, the new
architecture of the system allows read and write operations by
application programs to be made directly to the input/output
devices. This eliminates the cumbersome multi-step software
processes invoked by prior art systems using the operating system
and trusted code for every input/output access. In order to
accomplish the process safely, the input/output architecture of the
system 22 utilizes an input/output control unit 29 which provides
its own virtual name-to-physical-device and context translation for
all of the input/output devices associated with the new control
unit 29 on its own device bus 34. By enforcing this translation,
application programs can write directly to input/output devices
without affecting assets of other application programs. Once this
translation from virtual names furnished by the application
programs to physical input/output devices on the device bus is
accomplished and context has been furnished to the input/output
devices, translation of addresses of input/output devices on the
input/output bus into physical addresses on the device bus 34 is
accomplished directly by hardware at the input/output control unit
29. This hardware also checks permissions; and, when an operation
is known to be safe, it is performed by hardware. When a
translation operation fails, the operating system software is
invoked. Thus, rather than trapping every input/output operation to
determine whether it is safe as is done in prior art computer
systems based on the PDP11 architecture, the system 22 traps and
sends to operating system software only unsafe operations allowing
hardware to accomplish most translations and greatly speeding the
access of input/output devices.
The architecture of the system 22 has been designed so that it
eliminates almost all read operations of input/output devices by
the central processing unit. In order to accomplish this, the
input/output control unit 29 includes a first-in first-out (FIFO)
unit 31 for storing write operations directed to the input/output
control unit. The FIFO unit 31 queues incoming write operations.
Unlike FIFO units in prior art systems, it stores both addresses
and data. This allows the write operations to the input/output
control unit 29 to occur asynchronously so that both the central
processing unit 21 and the input/output control unit 29 may be
functioning independently of one another and neither need wait for
operations of the other.
To help maintain this asynchronous operating arrangement and to
eliminate read operations to the extent possible, the input/output
control unit also includes an advanced direct memory access (DMA)
device 35 which provides direct memory access for operations
conducted involving input/output devices. The DMA device 35 allows
the results of input/output operations to be written from
input/output devices to main memory 23 rather than requiring read
operations by the central processing unit 21 to obtain these
results. This eliminates almost all need for the central processing
unit 21 to read input/output devices and drastically increases the
overall speed of input/output operations. The DMA device 35
includes its own memory management unit which allows writes from
input/output devices to the virtual memory space of an application
without involving the operating system in the translation
process.
Although the input/output architecture of system 22 may be used
with systems utilizing a single input/output bus for all
operations, the preferred embodiment of system 22 functions as well
in a system utilizing a local bus 27 such as the Peripheral
Component Interconnect (PCI) bus or the Video Electronics Standards
Association (VESA) local bus which may be associated with other
input/output buses. While the discussion of this specification will
assume that bus 27 is a local bus, the local bus 27 is also
referred to in this specification as the input/output bus 27 in
order to emphasize its use. In arrangements utilizing local buses,
the central processing unit 21 and main memory 23 are typically
arranged on a processor bus 24 and a memory bus 26, respectively,
and are joined to a bridge unit 25. The central processing unit 21
typically includes a memory management unit. The bridge unit 25
provides write buffering for operations between the central
processing unit 21 and the input/output bus 27, between the central
processing unit 21 and main memory 23 on the processor bus 24 and
the memory bus 26, and between the input/output bus 27 and main
memory 23.
Typically, various input/output devices are arranged on the
input/output bus 27 as bus masters and bus slaves. In prior art
systems, these local bus masters and slaves are those components
(such as a graphics output device for connecting an output display
monitor or a hard disk controller unit) which require the most
rapid input/output operations for system success. If such local bus
masters and slaves are connected to the input/output bus 27, they
are utilized with the present architecture for the purpose of
running legacy programs and input/output functions not implemented
by the input/output control unit 29.
In the architecture of system 22, an input/output control unit 29
is shown joined to the input/output bus 27. The control unit 29
includes a hardware FIFO unit 31 for receiving incoming commands
addressed to the input/output devices on a device bus 34.
The general operation of the input/output unit 29: FIG. 2
illustrates the manner in which operations are conducted by
software in the new architecture. An application program which
utilizes the new architecture may issue a command requesting
permission from the operating system to map certain of the physical
addresses decoded by the input/output control unit 29 into the
address space of the application program. The operating system,
using a new I/O driver #1, allots some portion of the system
physical addresses which the input/output control unit 29 is
decoding to the particular application program address space for
its use only and installs the virtual-to-physical input/output bus
address translations for the application program in the memory
management unit. In a typical computer system, the memory
management unit stores translations for what are referred to as
pages of memory. If the size of the portion of system physical
addresses allotted to an application program is a multiple of the
memory management unit page size, then the I/O driver #1 can use
the memory management unit to ensure that no more than one
application program may access each area.
Installing the appropriate translations in the memory management
unit of the central processing unit 21 creates a path around the
operating system by which the application program may directly read
from and write to the hardware of the input/output control unit 29.
The application program then writes to these allotted input/output
bus addresses providing as data a virtual name of its choice for an
input/output device on the device bus 34. The input/output control
unit 29 takes the input/output address and the virtual name and
uses it to first create and then install a translation from
input/output bus addresses to device bus addresses in its internal
hardware and to place the context required by the application
program in that input/output device. Once this has occurred and for
so long as the application program continues to run, the
application program writes commands which the memory management
unit associated with the central processing unit translates to the
physical addresses on the input/output bus 27 for the input/output
control unit 29; and the input/output control unit 29 further
translates the input/output bus addresses of the commands to
physical addresses of input/output devices on the device bus 34. In
this way, the application may write directly to the input/output
unit in order to utilize an input/output device such as the
graphics output controller 33 without requiring any software
intervention by the operating system. As will be understood from
the more detailed description which follows, the use of many
identically-sized input/output device address spaces each assigned
for use only by one application program allows the input/output
addresses to be utilized to determine which application program has
initiated any particular input/output write operation.
Area addresses: In one embodiment, the FIFO unit 31 includes a
plurality of FIFO buffers 39 (see FIG. 3). When an application
program desires to write to an input/output device on the device
bus 34, it addresses that device. Decoding circuitry decodes the
address by reviewing a number of the highest order bits decoded by
the chip sufficient to indicate a unique portion of the
input/output address space assigned to an application program and
places the command in the appropriate FIFO buffer 39 for that
application program. Each FIFO buffer 39 handles commands only from
the application program to which the address area has been mapped.
An arrangement using a smaller number of FIFO buffers 39 than one
buffer for each application program (as few as a single buffer 39)
could also be utilized as long as the address of the FIFO buffer is
mapped to only one application program at a time. The unit 29
receives physical addresses furnished by the memory management unit
and virtual names furnished by application programs for operations
to be performed which have been sent to the FIFO unit 31 and
controls the translation of those virtual names for all
input/output devices. The hardware unit 29 includes the device bus
34 to which the individual input/output devices such as a disk
controller 32, a graphics output controller 33, and a sound
generator 37 are shown joined. The unit 29 also includes a DMA unit
35 which is adapted to transfer data between the individual
input/output devices and main memory for use by the central
processing unit or other components of the system.
Creation of a safe translation for an input/output device: When the
code of an application program is written to take advantage of the
new architecture, a safe translation for an input/output operation
utilizing a physical input/output device must first be created. A
safe translation for an application to utilize an input/output
device requires not only a correct physical address for the device
but also correct context so that the device will function
appropriately with the device. To create such a safe translation,
the application program sends a first special calling command
adapted to call an input/output device to the input/output control
unit 29; this special calling command includes as data a predefined
name such as "LINE.sub.-- DRAWER" selected by the application
program in accordance with a prescribed naming convention. The
command is transferred directly to the addressed one of the FIFO
buffers 39 of the FIFO unit 31 where it is placed in a FIFO queue.
At this point, the central processing unit 21 may go off to other
work. When this special calling command reaches the bottom of the
FIFO buffer 39, no translation between this predefined virtual name
and a physical address on the device bus 34 is resident in
hardware. This causes an interrupt, and the predefined name is sent
to a second new input/output driver called the "resource manager"
associated with the control unit 29. The resource manager keeps an
internal data base of data structures representing particular types
of input/output devices under the predefined names. The resource
manager looks up this known predefined name in its internal
database of data structures with predefined names and finds the
data structure defining that device in the data base. The resource
manager makes this predefined data structure available for
immediate use. In order to utilize the general device definition
provided by the data structure, the application program then
provides its own virtual name for that device as data and using a
"create" command, and the resource manager creates a new data
structure in its internal database using the virtual name the
application furnishes for that specific instance of the device
(e.g., MY.sub.-- LINE.sub.-- DRAWER). This new data structure
includes the various properties of the general device having the
data structure with the predefined name including the physical
address on the device bus 34 of the hardware which provides the
function for the predefined name and any context required by the
hardware for operation.
When the application program later wants to utilize that
newly-named object representing an input/output device, the
application program writes the virtual name chosen with a special
calling command which calls an object for the input/output device.
The resource manager looks up the new data structure which has been
created and (for a physical device) finds the context and physical
address on the device bus 34 for the particular input/output device
now described by the name. The resource manager changes any context
at the input/output control unit 29 required by the new
input/output device which has been named. The physical address on
the device bus 34 which has been found is then placed in hardware
to provide a translation from the input/output bus addresses so
that when commands are sent to the same input/output device from
the application program, they are routed by hardware to the
particular addressed input/output device on the device bus 34.
Unsafe operations: In any case in which the input/output device to
which the operation is directed is unknown to the control unit 29,
the unit 29 calls the "resource manager" which runs on the central
processing unit and functions as a portion of the operating system.
The resource manager determines how the operation is to be handled.
The operation may be a write by a new application program (such as
that described above) requiring various set up operations before it
may proceed. If an operation requires context changes at the
input/output device, this is handled by the resource manager. If an
operation requires a procedure which is not yet in order under the
operating system such as requiring data from memory which is not in
memory at that time, the resource manager transfers the command to
the operating system to perform the necessary memory transfers (or
the like) which allow the commanded operation to proceed.
Alternatively, the operation may be directed to a device which is
not otherwise associated with the control unit 29 such as a LAN
interface or other bus master or slave on the input/output bus 27
which is not manufactured to cooperate with the unit 29. If such a
device is addressed, the command is directed to the operating
system by the resource manager and handled by the operating system
in the normal manner for input/output devices of the prior art.
Address translations in hardware: When the operation involves a
device directly associated with the control unit 29 on its device
bus 34, the commands after the first commands (creating the new
data structure, attaching its new virtual name, providing any
necessary device context, and creating the address translation) are
sent by hardware directly to that device for execution. If the
command requires that data be transferred to or from the
application, the input/output device performs the transfer using
the DMA unit 35. Upon the return of data in response to a command,
the DMA unit 35 of the control unit 29 responds by transferring the
data to main memory and notifying the central processing unit in a
separate DMA operation of the existence of the data so that no
local bus read operation by the central processing unit 21 is
necessary to ascertain the result of the operation or to receive
the data provided.
Legacy operations: In contrast to the operation discussed above, if
an application program does not utilize the advantages of the new
input/output system, it may still function in the manner of
existing applications running on prior art systems. For example,
older application programs operating in a multitasking system which
have no knowledge of the new input/output system and are attempting
a subroutine call to request the operating system to perform an
operation using an input/output device associated with the unit 29
will trap into the operating system where its permission to proceed
will be checked. The operating system will translate the call to
the appropriate physical address and, finally, call the trusted
code of the new system I/O driver #1 to execute the command. The
new system I/O driver #1 functions in the manner of a typical
driver of the prior art and executes the command by writing from
its library of operations to the input/output control unit 29 in
the manner described above for application programs with knowledge
of the input/output control unit 29. In fact, the new I/O driver #1
functions in a manner essentially identical to an application
program with knowledge of the control unit 29 by providing a
virtual name for the device specified to which the physical
addresses for that device may be attached with a command calling
the device. The new driver #1 has mapped to its address space a
portion of the physical addresses decoded by the unit 29. The
command data generated in response to the command from the older
program is then transferred by this driver to the FIFO unit 31 and
processed in the same manner as are direct operations from an
application with knowledge of the unit 29. Although this new I/O
driver #1 functions as do other prior art drivers requiring the use
of the operating system and stepping through the various stages of
translation and permission checks, legacy software may utilize the
new architecture without any additional requirements being placed
on the system other than those which exist in the prior art.
Moreover, this legacy code will run faster than on prior art
systems because of the asynchronous result provided by the FIFO
unit 31 and the write only operations that unit supports.
Specific details of the new architecture: FIG. 3 is a block diagram
illustrating details of the input/output control unit 29 including
the device bus 34 and the input/output devices arranged on that
bus. As described above, the input/output control unit 29 includes
a decode unit 30 which receives commands directly from the
input/output bus 27 and transfers the commands to a pusher circuit
53 which transfers the command to one of the FIFO buffers 39 of the
FIFO unit 31. The FIFO unit 31 stores the data along with the
addresses for each of the commands being transferred to all of the
input/output devices associated with the input/output control unit
29. The buffers 39 of the FIFO unit replace the individual data
registers used by the input/output devices of the prior art.
However, in contrast to the registers used by the prior art for
communication on the input/output bus 27, the FIFO unit 31 allows a
very much larger number of commands to be processed much more
rapidly and facilitates the asynchronous operations of the
input/output devices and the central processing unit. In one
embodiment, the FIFO unit 31 includes 128 individual FIFO buffers
39 each of which has thirty-two stages. The use of 128 individual
FIFO buffers 39 allows a FIFO buffer to be assigned to each of 128
different application programs which may be running on the central
processing unit for the use of that application program alone. The
thirty-two stages of each buffer 39 allow any FIFO buffer to hold
thirty-two individual serially ordered commands at any time.
Although each of the stages of the FIFO unit 31 holds (along with
the address bits) the data held by an individual register of a
typical prior art input/output device, a particular system has the
ability to store commands for over 16 thousand 32 bit registers for
each of 128 different application programs which may map different
addresses decoded by the input/output control unit 29.
Addresses: In one embodiment, the entries in each FIFO buffer 39
include thirty-two bits of data space and sixteen bits of the
twenty-three bits of address space decoded by the input/output
control unit 29. The remaining upper seven bits of the address
represent the 128 distinct areas which are available and thus
define the particular one of the FIFO buffers 39 which is utilized
by a particular application program. The address and data space of
the total command space (including that in the FIFO unit and the
seven highest order bits defining the area assigned to the
application program by the operating system) are pictured in FIG.
4. The twenty-three bits of address space are sufficient to map
eight megabytes of address space on the input/output control unit
29. The eight megabytes of address space is divided into the 128
individual areas each having 64 Kbytes which may be allotted by the
operating system to an application program.
When a first command from an application program is transferred to
the input/output control unit 29, decoding circuitry selects one of
the FIFO buffers 39 using the upper seven bits of the twenty-three
address bits representing the area designated for that program and
transfers the command to an input stage of that FIFO buffer 39.
Each of the 128 addressable areas is subdivided into eight separate
sub-areas each having eight Kbytes of address space. The next lower
three bits of the address space represent these eight sub-areas.
The application treats each of these eight sub-areas identically,
designating at various times various objects representing
particular input/output devices and their context, as being
accessed through each sub-area. As will be seen later, each of
these sub-area addresses represents one of eight registers which
may store the physical address of an input/output device on the bus
34. The two lowest order bits of the address space represent byte
positions in a command. In the preferred embodiment, the data is
word aligned; and these bits are not included in the FIFO buffer
39. Consequently, eleven bits are left to represent a particular
operation using the particular input/output device designated by
the sub-area. With eleven bits of address space, 2048 individual
operations (or portions thereof) are available for devices using
addresses in each sub-area. In one embodiment, data structures
(objects) are created in an object oriented programming language to
represent the devices and their contexts which may be addressed in
the sub-areas. The operations of the devices are then encoded as
methods on each of these objects. This encoding of a sub-area as an
object of a particular class is dynamic, and a new object
representing a new device and its context may be encoded in the
sub-area by an application program writing to offset zero, the
special calling command which calls an address translation for a
new device to the sub-area holding the address translation of an
old object.
As pointed out above, when a program which is able to utilize the
input/output system first requests that the operating system map a
portion of the addresses decoded by the input/output control unit
29 to its address space, the operating system assigns physical
addresses designating one of the 128 areas available for the
input/output control unit 29 to the application. Since the I/O
driver #1 is constructed never to map more than one application
program to an area, the seven bits also identify the application
program and the FIFO buffer 39 which it utilizes. When an
application program writes to the FIFO unit 31, the seven upper
bits of the address are used to determine the sixty-four Kbyte area
which it has been allotted by the operating system and the FIFO
buffer 39 to which it is assigned. The three bit sub-area
designation is used to select one of the eight Kbyte sub-areas
which it may allot to a device. The eleven bit offset is used to
determine the method or operation to be carried out by the device,
and the thirty-two bit data space is used for the data related to
the commanded operation. In a typical write operation, the write to
any particular eleven bit offset invokes a particular method
(operation to be performed indicated by the eleven bits) on the
object (the input/output asset designated by the present name for
the sub-area). However, these bits are also interpreted (1) to
create a new data structure representing input/output devices which
respond to virtual names given by the application program, (2) to
provide direct translations from virtual names to physical
addresses of input/output devices on the device bus 34, and (3) to
call the resource manager to perform various software
operations.
The puller circuit, current address registers, and translation
table: These operations are accomplished by various circuitry and
the resource manager, particularly by a puller circuit 40, a
current physical address table 36 which includes eight address
translations for devices presently in use, and a translation table
38 which may include a much lager number of address translations.
In order to correctly direct the address and data provided in each
command, each FIFO buffer 39 includes a circuit 40 called a puller
which reviews the command about to be executed. The puller circuit
is illustrated in more detail in FIG. 7. The puller circuit 40
looks at the address of the bottom command in the FIFO buffer 39.
The puller circuit 40 uses the three sub-area bits of the address
to determine which of the eight sub-areas (current address
registers) of the table 36 associated with that buffer 39 is to
receive the result of a lookup in the translation table 38.
Writes to zero offset: The puller circuit 40 also includes logic 71
which then determines whether the next eleven method bits of the
address are all zero. If the eleven method bits are all zero, this
indicates a write to the zero offset which is the special calling
method used for indicating that an application wants a new
translation for an input/output device; and the puller circuit 40
sends the data to the translation table 38 along with the upper
seven bits from the address indicating the area and performs a
lookup. It will be recalled that when the write is to this special
calling method, the data is the virtual name of a device. The
result of the lookup is an address on the device bus 34 and an
instance number defining context which are placed into one of eight
registers of the table 36 referenced by the three bit sub-area.
When the physical address and instance number are placed in the
register of the table 36, the puller circuit sends the seven bits
indicating the area and the instance number to the input/output
device to change the context on the device. This is the manner in
which translations are made available for immediate use.
Writes to non-zero offset: If the offset is not zero, the puller
circuit 40 takes the three bits indicating the sub-area and indexes
into the table 36 to the proper register to find the device bus
physical address. The puller circuit 40 concatenates that address
with the eleven bit offset designating the method and writes the
method and thirty-two bits of data to that physical address on the
bus 34. However, if the value read from the sub-area of the table
36 is a special value which indicates a failed translation, this
value generates an interrupt which calls the resource manager. The
resource manager then uses the command at the bottom of the FIFO
buffer 39 to perform whatever software operation is required by the
command. This helps assure that unsafe operations are handled by
the operating system.
FIG. 5 illustrates in the first two lines one entry in the
translation table 38 utilized in one embodiment of the new
architecture. As may be seen, the seven bits of the address
designating the area assigned to an application program and the
thirty-two bit virtual name translate into twenty-three bits, seven
of which indicate the address of the physical device on the device
bus and sixteen of which indicate the instance of the data
structure which provides the context to be placed on the
input/output device. Additional control bits may also be included
as a part of the translation data stored in the table 38 but are
not shown. Each of the last two lines of FIG. 5 indicates one way
in which the bits obtained in the translation are used. The eleven
bits indicating the method invoked are concatenated with the
physical address for the device retrieved from the translation
table 38, and the concatenated value is placed on the bus 34 with
data. Each of the input/output devices decodes addresses on the bus
34 to determine if it is the addressed device and responds
accordingly to the operation indicated by the method.
Creation of a data structure: When an application program first
writes to the area which it has been allotted by the operating
system and is assigned to a FIFO buffer 39, the command is
ultimately reviewed by the associated puller circuit 40. The puller
circuit will find that the application program has selected one of
the sub-areas using the three bit sub-area designation, selected an
offset zero using the eleven bits, and has written a predefined
name for a particular input/output device in the thirty-two bit
data space. When the application program selects a zero offset as
the eleven bits representing an operation, the application is
indicating that it desires to call a data structure which has been
named and make it immediately available for use. When a zero value
is written as the eleven bit offset to any one of the sub-areas,
this instructs the input/output control unit 29 to make available
that one of the sub-areas to the newly-named object and to
interpret eleven bit offsets within the sub-area as the various
methods which are available to an object of that class.
When the application program writes the name of an object as data
to offset zero of a sub-area, the puller circuit 40 takes the
virtual name, adds the seven bits designating the area, and looks
up the concatenated value in the translation table 38 to obtain the
physical address on the device bus 34 and the instance number of
the physical device which is responsible for the operation
represented by the particular object being named. If a translation
is in the table 38 and the object represents a line drawing device,
the physical address on the bus 34 of the line drawing hardware in
the graphics rendering engine should be returned. When the physical
address is returned, it is placed in one of eight positions
(registers) of the current physical address table 36 designating
the sub-area to which the zero offset was written. If the
translation for the physical object does not exist in the
translation table 38 of the input/output control unit 29, however,
the input/output control unit returns a miss. This transfers the
operation to the resource manager. The resource manager places a
special value (all zeros in one embodiment) in the appropriate
register of the table 36 and uses the command at the bottom of the
FIFO buffer to perform whatever software operation is required by
the command.
On a first write to the input/output control unit 29 by an
application program, there will be no translations for that object
name in the translation table; and the operation will be
transferred to the resource manager. The resource manager in the
preferred embodiment of the new architecture has access to the
database which includes the data structures for a number of
predefined objects. These objects may represent hardware or
software which implements various portions of the input/output
operations. When an application program writes the name of a
predefined object at an offset zero in one of the eight sub-areas,
this is a request to the resource manager to make the predefined
object one of the eight objects available for immediate use in one
of the eight sub-areas. The application program follows this
command with a command directed to the same sub-area to create an
instance of the predefined object in the database and name it as
the application program defines in the data bits of the
command.
The resource manager reviews the details of the command being
written and determines that is a write to a zero offset. This
causes the resource manager to look at the predefined name to
determine the class of the object. When it determines that this is
a name for one of the predefined general classes of objects
associated with the input/output control unit 29, the resource
manager looks up the data structure for that object and makes that
object immediately available. To make the object immediately
available, the resource manager allots the sub-area to the
predefined object but also places a special code in the table 36 to
indicate that the object is a software object and the resource
manager is to be called when the predefined object is addressed.
Then the resource manager interprets the create command which
follows as a create method for the predefined object and creates a
new instance of the predefined object, names the instance using the
name requested by the application program, and stores it as a new
data structure in the object database.
Modifying context of a newly-created data structure: When the
application program later desires to utilize the input/output
device for which it has created a new object, it writes the name it
has selected for the object as data to the zero offset address. The
puller circuit 40 causes a lookup in the translation table 38 using
the new virtual name and the seven bit area identification. Again,
there will be no translation for that virtual device name in the
translation table 38 since the data structure which has been
created is a software object which has no translation in the table
38; and the operation will be transferred to the resource manager.
The resource manager reviews the command and determines that is a
write to a zero offset. This causes the resource manager to look up
the new data structure with that virtual name in the object
database to find the object which defines the input/output device.
The resource manager uses the seven bits designating the area
allotted to the application program and the thirty-two data bits
providing the virtual name given by the application to find objects
in its database.
When the resource manager finds the data structure, it places the
special value in the addressed register of the table 36 instead of
an address on the device bus 34 to indicate that this is still a
software object. Until the physical device is utilized, the
application program may send various commands as methods on the new
object; and these will be executed by the resource manager. A
plurality of low numbered offsets are utilized for modification of
a software data structure. For example, the application program may
send commands which set the details of the appropriate context
values for that particular device functioning with the particular
application program for the particular purpose. This changing of
context from the context provided by the predefined data structure
typically occurs before the device is utilized while only the
software object is affected.
Placing safe translations in the translation table: Ultimately,
when a physical input/output device receives a command which makes
a first use of the physical device, the resource manager places a
translation for the particular virtual-name-to-device-bus-address
of the appropriate physical device in the translation table 38.
It should be noted that the virtual name selected by an application
program for a particular data structure representing an
input/output device and its context is used for the later retrieval
of the address translation for that input/output device. In fact, a
number of different application programs may use the same virtual
name for the same or different virtual objects without causing any
ambiguity, for each object created has its own separate area
address bits which relate to that application alone.
In any case in which a translation for the virtual name to the
device bus address for a new physical object is placed in the
translation table 38, a number of additional bits which define the
instance of the data structure and therefore indicate any context
which is presently a part of the data structure and is necessary
for the operation of the device with the application is also stored
in the translation table 38 by the resource manager. As will be
seen by doing this, the translation table 38 is being used to trap
operations which require context switching before a device is
allowed to perform an operation. Finally, the resource manager
restarts the write operation. The lookup in the translation table
38 then succeeds. This causes the physical address and instance
value to be placed in the register of the table 36 and the puller
40 to send the seven area bits and instance value to the
input/output device to change the device context.
When the physical address on the device bus 34 and the instance
value of the device corresponding to the current object are first
placed in a register of the current address table 36, the address
is used by the puller to send the instance value and the seven bits
indicating the application program (and the address area) to the
device on the device bus 34 (see line three of FIG. 5). The device
compares the seven bits and the instance value to the area and
instance it is presently utilizing. If they differ, the device
changes its context or interrupts the resource manager to change
its context so that the device is properly initialized for the
application program.
Thus, whenever an application program selects a different
input/output device to utilize a sub-area of the address space by
writing to offset zero of that sub-area, the change of input/output
device causes the puller to send the area bits and the instance
value to the input/output device to change any required
context.
When an application program writes the virtual name of an object to
offset zero in one of the sub-areas, and when the lookup in table
38 of that virtual name succeeds, the physical address of the
corresponding device on the device bus 34 and the instance value
are also stored in a slot of the eight entry current physical
address table 36 which slot corresponds to the sub-area to which
the virtual name was written. The table 36 stores the physical
address on the device bus 34 of the device corresponding to the
current object accessible in that sub-area, if there is such a
device. If there is not a physical device or there is no
translation in the table 38, the entry stores the special value
which has no translation and therefore causes the input/output
control unit 29 to interrupt into the resource manager.
Writing directly to input/output devices: After the physical
address on the device bus 34 of the device corresponding to the
current object has been placed in the current address table 36,
when a next write occurs to that object as indicated by the three
bits of the address selecting the particular sub-area, the offset
address will typically be other than zero. This offset will
indicate the method invoked on the object. This offset (indicated
by the eleven bits) is concatenated with the physical address held
in the table 36 (see line 4 of FIG. 5) and broadcast on the device
bus 34 to select the particular input/output device and the
operation indicated by the method which is to be performed by that
device. All of the devices on the device bus 34 listen on the bus
and decode commands addressed to them.
Current address registers and sub-areas: Since eight sub-areas are
available at once through the current address table 36, an
application program may write up to eight virtual names for devices
the application desires to utilize in input/output operations and
have physical addresses for those devices immediately available by
simply writing the virtual name to the zero offset of a sub-area.
Thus, up to eight objects (devices) may have address translations
immediately available in the table 36 for the application program
using the FIFO unit 31.
The eight sub-areas available provide a large number of output
options for an application program. The availability of eight
sub-areas allows the application to accomplish a number of
functions without the necessity of a translation table lookup and
thus speeds input/output operations. However, since any application
program may need to have access to all of the input/output assets
which are available, the system provides a rapid manner of
providing assets in addition to the eight devices which are
represented by objects which fill the eight sub-areas allotted to
that application program. When all of the eight sub-areas have been
used by an application program so that input/output-to-device bus
physical address translations for a device exist in each of the
eight spaces in the table 36 and the application program running
desires to write to a different input/output device, the
application program may select a new device which it desires to use
and place its address translation in the table 36 in place of any
address translation presently occupying one of the registers. To
accomplish this, the application program writes a new virtual name
of a device as data directed to the zero offset of any of the eight
sub-areas. This causes the input/output control unit 29 to replace
the object presently occupying the sub-area with a new object
representing the device indicated by the newly presented virtual
name. This is accomplished by the puller circuit 40 initiating a
lookup in the translation table 38 and a replacement of the
physical address in the table 36 designating the object in the
sub-area with the physical address of the new device if a
translation for the new object for the physical device has already
been placed in the translation table 38 by the resource manager.
Whenever an application program places a different translation in a
register of the table 36, the change of address causes the puller
to send the area bits and the instance value to the input/output
device to change any required context.
However, if this is the first use of this object by the application
program, the name-to-physical-address-translation is not in the
translation table 38. The new virtual name causes the
virtual-name-to-physical-address translation to miss in the
translation table 38 so that the operation is trapped and sent to
the resource manager. Presuming that an instance of a predefined
data structure has already been created under the virtual name, the
resource manager recognizes the zero offset as calling for a new
object, reviews the new name, and finds the data structure for that
name in the database. It uses the object data structure to obtain
the instance value indicating the context for that new input/output
device and writes the virtual-name-to-physical-address translation
and instance value into the translation table 38. The operation
then proceeds and succeeds, the physical address and instance value
for the object is placed in the current physical address table 36
in the register in which the object being replaced previously
resided, and the context of the device is updated. When the next
write occurs for that named input/output device, the physical
address translations for that device (object) will be in the
current physical address table 36 so that it may be immediately
placed on the bus 34. Thus, the resource manager is called and
assures that the context on an input/output device is correct
before its address translation is placed in the physical address
table 36.
Whenever any object is named for which the physical address is not
in the physical address table 36 but for which a translation is
available in the translation table 38, the lookup of that virtual
name succeeds, the physical address and instance number of the
corresponding device on the device bus 34 is stored in a slot of
the current physical address table which corresponds to the
sub-area to which the virtual name was written. Thereafter, writing
to an offset to this sub-area will indicate a method invoked on the
new object in the sub-area. This method (indicated by the eleven
bits) is concatenated with the physical address held in the table
36 and broadcast on the device bus 34 to select the particular
input/output device and the operation (indicated by the method)
which is to be performed by that device. In this manner, the tables
36 and 38 act as a two level cache for object name translations
which the application utilizing the FIFO unit 31 may immediately
access and makes an extraordinarily large number of operations
available even though the physical address space allotted to the
program is limited.
Although 2048 operations are available for each object which is
physically on the device bus 34, it is probable that some number of
the operations (methods) will not be implemented in hardware. When
an input/output device receives a command including a method it
cannot carry out, the device addressed responds to the command
indicated by the offset by generating an interrupt indicating that
the hardware cannot deal with the operation. The interrupt calls
the software of the resource manager so that the resource manager
may accomplish the operation. This allows those operations which
are invoked very infrequently to be carried out in software, while
those operations which are used frequently are implemented in
hardware in order to speed up the system. In order to assist this
operation, each input/output device on the device bus 34 also
provides a signal to the puller circuit 40 to signal the puller
circuit that no commands are to transferred to the input/output
device which has generated the interrupt until the interrupt
servicing has been completed.
Thus, as may be seen, the resource manager is a piece of software
which is associated with the input/output control unit 29 and
determines that the input/output control unit 29 functions
correctly. It maintains a database of data structures which
represent the various input/output devices and the context that
those devices require to function correctly. It fills the
translation table 38, does the necessary context switching for
initializing the physical devices, provides routines for less used
input/output operations which input/output devices may invoke
through interrupts, and does other things required to run the
input/output control unit 29. The resource manager may be thought
of as part of the operating system and takes the place of the
device driver used in a conventional input/output system. The
resource manager maps in a part of the physical hardware of the
input/output control unit 29 called the privileged address space.
This space is distinct from the FIFO unit. Unlike the application
programs operating with input/output devices, the resource manager
both reads and writes this address space to perform its various
tasks such as context switching. Unlike all of the device drivers
of the prior art, the resource manager accomplishes its functions
after the hardware of the input/output control unit 29 has been
directly addressed by an application program rather than before.
Moreover, in the overall operation of the input/output control unit
29, the resource manager is used infrequently compared to the
hardware portions of the input/output control unit 29 since the
resource manager attends only to creation operations, the various
software implementations of methods, and unsafe operations.
When a first application program writes to the input/output control
unit 29 in the embodiment illustrated, one of the FIFO buffers 39
of the FIFO unit 31 is dedicated to its use depending on the seven
bits designating the area mapped. The various commands written will
gradually fill the FIFO buffer 39. Since an area is mapped to the
address space of only one application program, all of the commands
in the FIFO buffer 39 are directed to responding to that program.
Moreover, once an object has been made accessible in a sub-area,
the three bits designating that sub-area indicate the meaning of
the eleven bits designating the operations which apply to the
object in that sub-area. When a new object is made accessible in a
sub-area by writing to a zero offset for that sub-area, commands to
that sub-area call forth the particular methods (operations)
related to that object (device).
Because the amount of space in each of the FIFO buffers is limited
to thirty-two entries in the embodiment being described (and will
be limited to some finite number in any embodiment), it is possible
for a buffer 39 to fill with data and be unable to take more
commands directed to the particular area. As discussed previously,
this will occur when the central processing unit is writing a FIFO
buffer 39 faster than the input/output control unit 29 is able to
handle the commands placed in the FIFO unit 31. In such a case, the
input/output control unit 29 issues a hold command on the bus 27 to
stop additional commands from being sent by the application
program. During the holdoff period, the puller circuit 40 may empty
some of the commands of the application program from the FIFO
buffer 39 so that space will be available. Alternatively, when the
bus holdoff expires, commands from the program may still fill the
FIFO buffer 39; and the new commands may have to be otherwise dealt
with or the data will be lost to the input/output control unit
29.
In one embodiment of the new architecture, a flow control register
45 is included for each FIFO buffer 39 of the input/output control
unit 29. The register 45 stores an indication of the number of
available spaces for commands in the associated FIFO buffer 39. The
use of this flow control register 45 allows a requirement to be
placed on an application program attempting an access of the
input/output control unit 29 that it first determine whether space
is available in the FIFO unit 31 before it writes any command.
Before writing any data, the application program must determine the
amount of "free" space to which it may write. This it does by
having the central processing unit read the value in the flow
control register 45 associated with the FIFO buffer 39 being
utilized. Once this value has been read, the application program
may send up to that amount of data before it need test again to
determine whether free space remains to which it may send more
data.
A flow control register 45 may be very simply implemented by
utilizing a single bit in each stage of the FIFO buffer 39 to
provide an indication that the stage is empty or filled. For
example, a valid bit in each stage of the FIFO buffer 39 may act as
such a register 45. If so implemented, the count of empty stages
may be translated by a hardware or software algorithm to a binary
number to be used to respond to a read by the central processing
unit.
In one embodiment of the new architecture the individual buffers 39
are implemented utilizing an array of random access memory (RAM).
In such an arrangement, the data comprising the commands is placed
in the RAM and pointers are utilized to define the positions of the
beginning and end of the individual FIFO buffers 39 and the
beginning and end of the data. In such an embodiment, a simple
subtraction of a pointer designating the last data placed in a
particular RAM buffer from the pointer designating the end of the
buffer provides the value to be held by the flow control register.
In such a case, the flow control register itself may be a logical
register whose contents are computed whenever it is read.
The use of the flow control register 45 allows an application
program presently using the resources of the input/output control
unit 29 to proceed without causing the overflow of the FIFO unit
31. Whenever the application has written commands totaling the
amount previously read from the register 45, it must read the
register again. This is especially useful for application programs
which may wish to transfer a series of commands in sequence,
possibly using the burst transfer mode of operation provided by
many of the modern buses. By ascertaining the space available in
the FIFO buffer 39 before transferring a sequence of commands, the
application can know that sufficient space will be available for
the operation. Provided that no application program ever writes
more than the data for which it has permission, no data will be
lost.
Various operating conditions can cause the flow control register 45
associated with a buffer 39 to store different values. If the
central processing unit is writing at a rate faster than the
input/output device can handle the commands, then a FIFO buffer 39
which is initially empty and provides a maximum free count value
when read by the central processing unit will provide a lower free
count number after the amount of data first indicated by the free
count register has been sent. This will occur because the FIFO
buffer 39 will empty more slowly than it is filled. In such a case,
the next value read by the application program after sending the
amount of data designated by the free count register will be a
smaller number. The application program may again send this amount
of data and be sure that the FIFO buffer will not overflow. On the
other hand, if the central processing unit is writing at a rate
slower than the input/output device can handle the commands, then a
FIFO buffer which is initially empty and provides a maximum free
count value when read by the central processing unit, will provide
the same free count number after the amount of data first indicated
by the flow control register has been sent. This will occur because
the FIFO will empty more rapidly than it is filled and will always
be empty when the register 45 is read by the central processing
unit.
One problem which must be appreciated is that the FIFO buffer
should be large enough to allow sufficient storage for the size of
the data to be transferred by any of the commands or the FIFO
buffer will never provide a free count value large enough for the
central processing unit to send an initial sequence of commands.
Thus, a FIFO buffer which is able to hold only 48 bytes of data
will never provide a free count large enough to allow the transfer
of data with a command which desires to transfer 64 bytes of
data.
In utilizing the free count value, the application program should
deduct the value of any data which it has already written before
each new write operation commences. In this manner, the application
program knows at any time the amount of data which it may still
write to the input/output device before another read of the free
count register will be required. This allows a number of write
operations to occur before a read of the register 45 (which
significantly slows operation of the overall system) need occur.
This also assures that any writes which might be presently in the
write channel on the input/output bus are accounted for in
determining the actual value of the free count left. It should be
noted that this read of the free count register is the only read
necessary in dealing with the input/output control unit 29. It
would be possible either to interrupt the central processing unit
if a FIFO buffer 39 filled or to use the DMA unit 35 to transfer
the value in the free count register to the central processing unit
instead of reading the value, but in the particular embodiment
reading the free count register appears to be the fastest
operation.
So long as there is at least one FIFO buffer 39 for each
application program which functions as a fixed allocation of local
storage for each input/output device, the assignment of another
FIFO buffer to another program does not reduce the amount of
storage available at the input/output device for the first program.
Each FIFO buffer acts as storage for the input/output device for
the particular program writing to the input/output device so that
space allotted to one program is still available to that program
even though another FIFO buffer is allotted to another program by
the input/output device.
In one embodiment, the application program may obtain a free count
value indicating the number of free entries in a FIFO buffer 39 by
reading from a designated method offset in any of the sub-areas of
the area mapped to the application program. The value read is that
stored in the flow control register 45 associated with the
particular FIFO buffer 39. This value is then placed on the bus
when the central processing unit provides the appropriate control
signals to implement the read operation. The application may write
up to the amount of data designated by this free count value
without further testing and be certain that overflow of the FIFO
buffer 39 will not occur.
As has been mentioned above, the central processing unit and the
FIFO buffers 39 do not normally operate at the same speeds. For
many hardware operations at the input/output device such as
describing the points of a polyline using a graphics controller,
the FIFO buffers 39 are able to operate much more rapidly than the
central processing unit is able to write data to the input/output
control unit 29. On the other hand, when the input/output control
unit 29 uses the resource manager to execute one of the commands,
the operation of the FIFO unit 31 slows down; and the central
processing unit may fill a FIFO buffer 39 much more rapidly than it
can be emptied.
It is very desirable to reduce the number of times the free count
register of a FIFO buffer 39 is read. This is true because read
operations take much longer than write operations. Before any read
operation can be carried out, all of the write operations in the
write pipeline from the central processing unit to the input/output
control unit 29 must be completed so that the write pipeline is
flushed. Reducing the number of free count register reads
significantly increases overall input/output system speed.
If a FIFO buffer 39 is emptying faster than it is being filled,
then each time the application program reads the associated free
count register to determine how much space there is left in the
FIFO buffer, the number read in the register will be gradually
increasing until it gets to the size of the FIFO and can get no
bigger. Thus, each time the application program reads the free
count register during such a situation, the read is actually a
waste. In fact, if a FIFO buffer 39 is actually emptying faster
than it is being filled, then the amount read is less and often
much less than the amount of data which can be sent before a next
read of the register is actually necessary.
In order to take advantage of situations in which a FIFO buffer 39
is actually emptying faster than it is being filled, an improved
design for the basic free count register arrangement has been
devised. The improved design provides an arrangement by which the
input/output control unit 29 may provide a larger number for the
free count register than the space actually available in the
associated FIFO buffer 39. This larger number when read by the
application program allows the application program to send a larger
number of commands to the input/output control unit 29 before a
read of the free count register is necessary. This reduces the
number of read operations of the free count register required by
the central processing unit and speeds the operation of the system.
In one embodiment, the improvement causes a reduction of total flow
control overhead on the bus 27 from fourteen percent of all bus
cycles to eight percent of the cycles and an attendant increase in
speed on the bus 27.
In the embodiment described which includes FIFO buffers 39 each
capable holding thirty-two commands each of which is thirty-two
bits in length, a FIFO buffer holds 128 bytes of data.
Consequently, a central processing unit read operation would
normally be required after at most 128 bytes has been written by an
application program. Using the described technique, the number in
the free count register may be increased to values up to an amount
equal to the amount of space available in a FIFO buffer 39 and a
runout storage area. In one embodiment, the runout area may be set
to hold 1024, 2048, 4096, or 8192 bytes of data. This gives the
input/output control unit 29 the ability to provide a number of
different free count values depending upon the particular
application program which is running. The actual free count value
may be set to such fixed increments, or an unlimited variation in
the free count value subject to the amount of runout space
available could be easily implemented.
It will be noted that reading the free count register is, in
effect, a space allocation to the application program promising
essentially that this much space is available and will not be taken
away. Since the input/output control unit 29 is not keeping this
promise when greater numbers are placed in the free count register
than space is available, the practice of using a larger value than
the space available is referred to as "lying."
Under optimal conditions, the process of lying works well and
allows the system to run more rapidly whenever the hardware of the
input/output control unit 29 is functioning more rapidly than the
central processing unit can fill the FIFO buffers. However, there
are times when the input/output control unit 29 slows down while
the central processing unit speeds up so that the input/output
control unit 29 may be caught lying. Under less than optimal
conditions, the input/output device will receive a command which
slows its operation; and it will begin to function more slowly than
writes are being presented by the central processing unit. In such
a case, the central processing unit will be filling a FIFO buffer
39 faster than the input/output control unit 29 can empty it. Since
the input/output control unit 29 has furnished a value greater than
the size of space available in the FIFO buffer to the application
program from the free count register and the application program
may write this larger amount of data before again checking the
value in the free count register, there must be some arrangement to
assure that data will not be lost. The arrangement described
provides a runout area of memory to which data may be written when
the input/output control unit 29 is caught lying. In one
embodiment, this may be a part of the memory used by a graphics
display controller which is in excess of the amount needed for the
purpose of a frame buffer. In other embodiments, other types of
memory which are associated with the input/output control unit 29
and may be conveniently accessed by the resource manager may be
used. However, the size of the memory must be sufficient to provide
space for data directed to all of the FIFO buffers which may
overflow if caught lying.
In a case in which an application program has been given a value
larger than the amount of space available in the FIFO buffer 39 to
which the application program is writing and the input/output
control unit 29 receives a command when the FIFO buffer is full,
the pusher 53 generates a holdoff command to the bus 27 in order to
stop the flow of write commands as soon as possible. If after the
holdoff period has expired, the FIFO buffer is still full, the
pusher 53 sets the value in the free count register 45 to zero and
transfers each of the next commands received on the bus 27 to
runoff memory associated with the input/output control unit 29,
recording the address and data of each command as it is stored. The
pusher then interrupts the resource manager which waits for the
puller circuit 40 to complete processing the commands remaining in
the FIFO buffer. When the puller circuit 40 has completed emptying
the FIFO buffer 39, the resource manager takes over the operation
of processing the commands.
The resource manager takes over each of the functions of the puller
in so far as the transferring of commands to the various portions
of the input/output control unit 29 are concerned until any runout
memory storing commands originally directed to that FIFO buffer 39
are empty. The resource manager must execute the commands in
sequence in order for the operations to be executed correctly. For
this reason, all commands addressed to the overflowing FIFO buffer
39 after a first command is sent to the runout area must also be
sent to the runout area. The value in the free count register is
held at zero until all of the data in both the FIFO buffer 39 and
the runout area have been cleared.
The pusher circuit 53 shown in detail in FIG. 7 is designed to
assist in controlling the operations related to FIFO changes
including changing application programs using a FIFO buffer 39 and
controlling overflow. With each new command being transferred to a
FIFO buffer 39, logic circuitry 73 matches the area identification
bits with those of the incoming application and detects whether a
different application has commands in the FIFO buffer 39. If the
seven area bits of the commands in the FIFO buffer do not match
those bits of the new command, the command is sent to the runout
area of memory. If the commands in the buffer are those of the
application sending the new command, the command is sent to a logic
circuit 75 which determines whether commands have already
overflowed the FIFO buffer 39 and been sent to the runout memory.
If so, the new command is sent to the runout area as well. If not,
the command proceeds toward the FIFO buffer. Finally, the command
is sent to logic circuit 77 which determines whether there is space
available in the FIFO buffer by testing the actual count in the
free count register. If space exists, the command is placed in the
FIFO buffer. If no space exists, the command is sent to the runout
area and the fact of overflow is recorded in a register 78.
The commands including the addresses and data must be stored in the
runout area in accordance with their area addresses. This allows a
number of different FIFO buffers 39 to overflow at the same time
without disturbing the data being transferred by any of the
application programs. After the FIFO buffer has been emptied and
the commands in the runout area have been executed by the resource
manager, the resource manager may turn the puller back on, signal
the pusher 53 to allow the FIFO buffer 39 to fill again, reset the
value in the free count register, and allow the normal operation of
the input/output control unit 29 to continue.
In addition to handling the duties of the puller circuit 40, in one
embodiment, the resource manager also attends to executing in
sequence each of the commands in the runout memory assigned to that
buffer 39. This means that the resource manager is able to emulate
many of the operations which may be directed to all of the
input/output devices associated with the input/output control unit
29 on the bus 34. Since the operations of the input/output devices
are well known to those skilled in the art, the details of the
particular operations which are implemented in software are not
discussed here. However, it will be understood that the resource
manager should have software commands for performing some of the
operations of a graphics output controller, of a sound controller,
of a disk controller, and any other input/output device which may
be connected to the device bus 34. The resource manager may in many
instances hand the individual commands off to the hardware joined
to the input/output control unit 29 for execution rather than
executing the commands in software; this will occur in instances in
which the operation may be handled more expeditiously in
hardware.
In order to accomplish lying, the arrangement stores data which
indicates whether the device should allow lying or not in the data
structure defining each predefined object. When an instance of the
predefined object is created and given a virtual name by an
application program, this data is transferred to the new data
structure. In one embodiment, the data determining whether lying is
allowed is included as a single bit with the physical address of
the device. This bit is transferred to the translation table 38
when the object is called for use by the resource manager and is
placed in one of the eight current address registers of the table
36 when the object is placed in one of the sub-areas. If the bit is
in one condition, the free count register may be set to a value in
excess of the value of space actually available in the FIFO buffer
39. Typically, the bit will only be set for devices which conduct
operations which will drain the FIFO buffers rapidly. Since the
lying bit is stored with each object, the resource manager can
restrict lying to one or a few applications in order to conserve
runout memory. It should be noted that since software operation are
always slower than hardware operations, operations implemented in
software typically do not utilize lying.
Those skilled in the art will recognize that the use of 128
individual FIFO buffers 39 to receive commands written by the
central processing unit is very expensive and probably not
economically feasible for use with personal computers. In addition,
it is expensive of system resources for an operating system to
context switch the central processing unit from one application
program to another. Thus, the operating system will try to ensure
that once an application is given the central processing unit the
application will retain control long enough to amortize the cost of
the context switch. This means that typically in a one processor
system at most one FIFO buffer will be active at a time. In a
multi-processor system, typically at most as many FIFO buffers will
be active as there are processors.
One embodiment of the new architecture provides a means for
attaining almost the same speed as is possible with 128 FIFO
buffers but much more economically. Instead of 128 individual FIFO
buffers 39, the described embodiment utilizes a smaller number of
FIFO buffers 39 than the number of areas into which the
input/output address space may be divided and thus the number of
application programs which may be run. Utilizing a smaller number
of FIFO buffers causes each FIFO buffer to function as a cache for
commands which may be sent by any of the application programs. In
the preferred embodiment, one FIFO buffer 39 is used for each
processor which may be writing to the input/output control unit 29.
For example, if the computer utilizes two processors arranged as
symmetrical multi-processors, then two individual FIFO buffers are
utilized. If, as is usually the case, only a single central
processing unit is used, then a single FIFO buffer 39 may in the
limit act as a cache for commands from all application programs
which may be running on the central processing unit. Such an
arrangement is illustrated in FIG. 6. Of course, in either case
more FIFO buffers might be used.
In order to act as a FIFO cache, any FIFO buffer 39 stores commands
from only one application program at a time. To accomplish this,
the input/output control unit 29 also includes a register 42
associated with each FIFO buffer 39. The register 42 stores the
upper seven bits of the address designating the area (and thus the
application program) for the commands presently in the FIFO buffer
39. When a first command from an application program is transferred
to the input/output control unit 29 having an empty FIFO buffer 39,
the seven bits representing the input/output address area allotted
to that program are placed in the register 42 associated with that
FIFO buffer 39 where they are held until the application program
using the FIFO buffer 39 changes. Since each area is assigned to
only a single application program, these seven bits clearly
designate the particular one of the application programs presently
having access to the particular FIFO buffer 39. The bits in this
register are used to access the translation table 38 and the
current physical address table 36 to obtain translations from
virtual names to physical addresses in the manner explained
above.
As with the embodiments which use one FIFO buffer for each
application area, a free count register 45 and a puller are
included with each FIFO buffer 39 as a part of the input/output
control unit 29. The register 45 stores an indication of the number
of available entries in each FIFO buffer 39 of the FIFO unit 31 and
allows a requirement to be placed on an application program
attempting to access the input/output control unit 29 that it first
determine whether space is available in the FIFO unit 31. It also
allows an application presently using the resources of the
input/output control unit 29 to proceed without overflowing the
FIFO unit 31. The application program may obtain a boundary value
indicating the number of free entries in the FIFO buffer by reading
from a designated offset in any of the sub-areas of its mapped
area. The application may write up to the amount of data designated
by this boundary value without further testing and be certain that
overflow of the FIFO unit will not occur. As with the multiple FIFO
buffer embodiments, the FIFO buffer arrangements of the described
embodiment make use of a memory runout area; such a memory runout
area assures that a FIFO buffer may be used to cache commands from
a plurality of application programs. The use of a memory runout
area also allows the system to make use of lying in order to
enhance the speed of transfer by reducing the number of read
operations by the central processing unit of the free count
register 45.
In order to assure that the commands from different programs may
utilize a small number of FIFO buffers 39 in the manner of a cache,
the seven bits of each new command indicating the address area are
compared with the seven bits held in the register 42. If these bits
match and there is space in the associated FIFO buffer, the new
command is entered in the FIFO buffer. If the bits match and the
FIFO buffer is full, the bus will be held off; if the hold off
expires and the FIFO buffer is still full, the commands will be
stored in the runout area. If the bits match and commands for that
FIFO buffer are already in the runout area, the new commands will
be stored in the runout area. If the bits differ from those in the
register but a cache is empty and there are no commands in the
runout area, the seven bits for that command are written to the
register 42 associated with the empty FIFO buffer and the commands
are stored in that FIFO buffer. If the bits differ from those in
the register and the runout area is empty but none of the FIFO
buffers are empty, the new commands are stored in the runout area.
If the bits differ from those in the register and the runout area
is not empty, the new commands are stored in the runout area.
In order to keep the various application programs from overrunning
the runout area, each application which has data is the runout area
but none in a FIFO buffer will receive a zero value when it reads a
free count register to determine how much space it has available.
Once an application program has written commands which overflow the
FIFO buffer and are sent to the runout area, all additional write
must go to the runout area to maintain proper sequence. Once an
application program has written commands which overflow the FIFO
buffer and are sent to the runout area, the application program
will receive a zero value when it reads a free count register to
determine how much space it has available.
Thus, when in the embodiment illustrated in FIG. 6, an application
program writes to the input/output control unit 29, the FIFO buffer
39 is dedicated to its use. The various commands written by that
application program will gradually fill the FIFO unit. Since the
input/output control unit 29 knows that it is responding to
commands from a single application program (a single area is mapped
to the address space of the application program), all of the
commands in the FIFO buffer are directed to responding to that
program.
When, in the embodiment illustrated in FIG. 6, the input/output
control unit 29 is responding to commands from one application
program and receives a command from a second application program,
the FIFO buffer 39 changes and links itself to the second program.
This occurs under control of the pusher circuit 53. If the FIFO
buffer 39 is empty, this may occur immediately. If the FIFO buffer
is filled with commands from another application program, these
commands will be executed before the commands of the new
application program are handled. In some cases, this may require
that the commands from the second program be written into local
memory associated with the input/output control unit 29 so that
their execution will be delayed until the commands from the first
application program have cleared the FIFO buffer. In this way, the
FIFO buffer 39 provides the appearance of an 128 individual FIFO
buffers for incoming programs in this particular embodiment.
When a command from an application program written to an offset
zero with the virtual name of a device to make an input/output
object which has been created accessible in one of the sub-areas,
the virtual name of the device that corresponds to the object is
concatenated with the seven highest bits in the register 42
indicating the application area and looked up in the translation
table 38 to determine the physical address on the device bus 34 for
that object. If the translation exists in the hash table 38, the
physical address on the bus 34 is placed in the current physical
address table 36. The physical address on the bus may then be
concatenated with the eleven bits of the method designated in later
commands which will cause immediate transfer of the command on the
bus 34 to the device. The other operations of the input/output
control unit 29 using the embodiment with a single FIFO buffer
proceed in the manner described above for input/output control
units 29 which utilize a FIFO buffer for each application area.
Although the present invention has been described in terms of a
preferred embodiment, it will be appreciated that various
modifications and alterations might be made by those skilled in the
art without departing from the spirit and scope of the invention.
The invention should therefore be measured in terms of the claims
which follow.
* * * * *